Subscribe to eTOC

A Mouse Model Offers New Clues to the Underlying Mechanism at Work in Stuttering

Article In Brief

In a mouse model of stuttering, researchers found that loss of astrocytes seemed to lead to vocalization deficits associated with the speech disorder.


In a mouse model of stuttering (lower panel), there are fewer astrocytes, shown in green, compared with controls (upper panel) in the corpus callosum, the area of the brain that enables the left and right hemispheres to communicate.

Loss of astrocytes seems to contribute to abnormal vocal patterns in mice genetically engineered with variants associated with persistent developmental stuttering in humans, researchers found. The findings offer insights into the anatomic, cellular, and molecular deficits associated with the neurodevelopmental speech disorder, the study authors suggested.

“It's the first time that we have identified astrocytes as a potential source of the stuttering deficit. I expected to find neuronal cells in other brain regions such as the cerebellum, based on previous evidence that stuttering involves abnormalities in speech motor control,” said Dennis Drayna, PhD, scientist emeritus in the section on genetics of communication disorders at the National Institute on Deafness and Other Communication Disorders at the National Institutes of Health, who is the lead investigator of the study published in the August 20 online issue of Proceedings of the National Academy of Sciences.

Previous human brain imaging studies suggested the source of stuttering deficits could be in the white matter of the brain, including the corpus callosum where the astrocyte loss occurred. However, Dr. Drayna explained, those studies had insufficient resolution to detect subtle differences at the cellular or molecular level.

Study Design, Results

Dr. Drayna and his colleagues previously identified three mutations in the GNPTAB gene associated with stuttering. In this study, the researchers wanted to determine the effects of the gene mutations on mice vocalization and investigate the neuropathology of stuttering in the genetically engineered mice.

The researchers engineered mice with the GNPTAB stuttering mutations, including Ser321Gly (S321G) and Ala455Ser (A455S), and compared the mutant mice to wild-type littermates on day eight after birth. They recorded their vocal patterns—comprising interbout, or longer pauses between groups of consecutive syllables, and shorter, or intrabout pauses, between vocalized syllables—similar to those found in people who stutter.

The investigators calculated and compared the vocal patterns in the mutant and wild-type groups. They found that the S321G and A455S mice had significant increases in interbout pause durations compared with their wild-type littermates when the syllable cut-off was >50 syllables (p=0.040; p=0.0027, respectively). However, only the mice with the S321G mutation had significantly longer interbout pauses than those of wild-type littermates when the syllable cut-off was >15 (p = 0.00056).

In contrast to previous studies, the S321G mutant mice also showed a small but significant increase in the duration of intrabout pauses (p=0.010 for number of syllables> 15).

The S321G mice also had more repetition in their vocalization and gaps in the timing of their calls than their wild-type littermates, said Dr. Drayna.

Based on these findings, the researchers used only the S321G mice to study the neuropathology of stuttering caused by this mutation. Specifically, they looked at the astrocytes, oligodendrocytes, and microglial cells using immunohistochemistry with cell type-specific antibodies.

Compared with their wild-litter mates, the mutant mice showed a significant loss of astrocytes, which was particularly prominent in the corpus callosum. Moreover, based on advanced MRI methods, the researchers found that the local volume of the corpus callosum was reduced despite normal diffusion tensor MRI values, which reinforces the finding that a defect exists in this brain region.

Follow-up experiments in which the GNPTAB human stuttering mutation was introduced into individual brain cell types—rather than the entire mouse—confirmed that the vocalization defect is specific to astrocytes. The mice did not have the stuttering-like vocalizations when the mutation was engineered into other types of brain cells.

The main study limitation was the use of an animal model to study human stuttering, which required engineering the mice with human stuttering gene mutations. Although researchers recreated salient features of the disorder, Dr. Drayna cautioned against interpreting the results too broadly given that the genetic mutations discovered so far explain only about 21 percent of stuttering cases. “By no means have we explained all stuttering mutations—we still have a long way to go,” he said.


“Its the first time that we have identified astrocytes as a potential source of the stuttering deficit. I expected to find neuronal cells in other brain regions such as the cerebellum, based on previous evidence that stuttering involves abnormalities in speech motor control.”—DR. DENNIS DRAYNA

Expert Commentary

“The study adds considerably to the literature on genetic and physiological mechanisms involved in sound and speech motor output. As to whether that coalesces into being a model for disruptive sound output in people who stutter, the researchers have not demonstrated that,” said David B. Rosenfield, MD, FAAN, chair in speech and language in neurology in the Stanley H. Appel department of neurology at Houston Methodist Hospital/Weill Cornell College of Medicine.

An animal model should meet several criteria for developmental stuttering, Dr. Rosenfield said. “Since stuttering in humans is nonvoluntary, the model should demonstrate that animals with phonatory disruptions do not want the abnormality because of negative social consequences, and that its location is not random.”

In addition, the speech-motor problem should only occur when the animals communicate and not during other oral-motor activities such as chewing, swallowing or biting. Finally, repetitive silences or lapses in animal production or sound during communication may be an adaptive strategy to the “blocks” associated with stuttering rather than with stuttering itself, noted Dr. Rosenfield.

He acknowledged that the silences could be related to a disruption in timing caused by the deficit in astrocytes. “But the fact that there is overlap with stuttering adaptive behavior such as the silences doesn't mean that that overlap is part of the underlying disturbance.”

“The study's value was that it showed a relationship between genetic mutations and stuttering. For example, learning about the long pauses (interbout) in the vocalization of genetically engineered mice indicates that there is a temporal programing defect in these mice,” said Kenneth M. Heilman MD, FAAN, distinguished professor emeritus in the department of neurology at the University of Florida College of Medicine.

However, the study did not clarify how the genetic mutation is causing this programing deficit. “It is also unclear how callosal dysfunction would induce stuttering. I don't think there's any research in humans showing that injury to the corpus callosum induces stuttering,” added Dr. Heilman.

If an impairment exists in the corpus callosum, it should affect communication between the right and left hemispheres, he said.

“For example, people with callosal dysfunction due to lesions would show such signs such as a left alien hand or left-hand apraxia. Instead, previous research shows that people who stutter have other types of problems such as a central control disorder that interferes with the temporal neuronal firing patterns that are required for rapid and dynamic speech processing.”

When people with developmental stuttering use an apparatus that delays auditory feedback, he continued, many will show a reduction of their stuttering. In addition, when people with normal speech patterns use delayed auditory feedback, many will develop some stuttering, he said.

Dr. Heilman also questioned whether mice are a valid model to study human stuttering given that rodents lack the hemispheric specialization found in humans. “People who stutter do it most often when they are using propositional speech (expressing ideas, thoughts, questions with the use of words that have to be assembled). Many do not stutter when singing, and performing series speech, such as counting. In most people, it is the left hemisphere that mediates propositional speech and the right that mediates singing and series speech.”


Dr. Drayna receives a salary from the NIH National Institute on Deafness and Other Communication Disorders and travel-related expenses from the Stuttering Foundation of America. Dr. Heilman has received royalties from books he has authored or edited. Dr. Rosenfield had no disclosures.

Link Up for More Information

• Han TU, Root J, Reyes LD, et al. Human GNPTAB stuttering mutations engineered into mice cause vocalization deficits and astrocyte pathology in the corpus callosum Proc Natl Acad Sci 2019:116 (35):17515–17524.
    • Raza MH, Mattera R, Morell R, et al. Association between rare variants in AP4E1, a component of intracellular trafficking, and persistent stuttering Am J Hum Genet 2015; 97(5):715–725.
      • Raza MH, Domingues CE, Webster R, et al. Mucolipidosis types II and III and non-syndromic stuttering are associated with different variants in the same genes Eur J Hum Genet 2016;24(4):529–534.